Quick Buzz Feed
Experts warn that AI-powered policing could expand surveillance and amplify bias.
Major companies like Axon, Motorola Solutions, TRULEO, Flock Safety, Clearview AI, and Mark43 are offering advanced AI-powered tools to law enforcement. These tools facilitate tasks such as searching body-worn camera footage, analyzing vast datasets, reviewing digital evidence, summarizing case information (e.g., Mark43's ReportAI and BriefAI), and identifying potential suspects through facial recognition. While these technologies aim to reduce administrative workload and enhance efficiency, experts caution about their inherent risks. Concerns include the potential for AI-generated outputs to contain errors, reflect biases embedded in underlying data, and lead officers to over-rely on the technology's recommendations. The concept of 'agentic policing,' where integrated AI systems generate investigative leads and connections, is particularly worrying as it could obscure the decision-making process and challenge traditional investigative methods. There have also been instances of misuse, such as officers accessing automated license plate readers for personal surveillance, prompting some communities to reconsider their contracts with such providers.
Law enforcement agencies across the country are at different stages of AI adoption, balancing potential benefits with inherent risks. For example, the Montgomery County Police Department in Maryland is cautiously exploring AI applications for non-emergency call handling, translation, transcription services, and report writing to improve efficiency and reduce administrative burdens. In Arkansas, officials are developing the Arkansas Criminal Intelligence Network, a centralized cloud platform intended to integrate data across various police agencies and support advanced AI analytics. Hawaii's Maui County Council has invested significantly in high-tech policing tools, including AI-supported cameras and drones for real-time monitoring and emergency response. Similarly, the Honolulu Police Department is piloting an Axon-developed generative AI feature that aids in drafting police reports by leveraging video and audio transcriptions from body-worn cameras, illustrating a growing trend toward technology-assisted public safety operations.
The increasing integration of AI into police operations, particularly in report writing and evidence analysis, raises substantial legal and evidentiary concerns. Errors introduced by AI systems, such as inaccuracies, omissions, or misinterpretations of context, could significantly impact how evidence is understood and used in investigations and court proceedings. To address these risks, experts and industry leaders advocate for critical safeguards: mandatory disclosure when AI is used in reports, rigorous human verification of all AI-generated text, regular independent auditing of these tools, and comprehensive training for law enforcement and legal professionals. This training should focus on how AI systems function and how to trace AI outputs back to raw source evidence. These recommendations align with frameworks, such as those from the Council on Criminal Justice, emphasizing independent validation, enforceable procurement standards, continuous performance monitoring, and clear human oversight to ensure that operators can override AI-generated outputs, highlighting the profound opportunities and challenges AI presents to criminal justice.